Chemical basis of the synergism and antagonism in microbial communities in the nests of

leaf-cutting ants
Author(s): Ilka Schoenian, Michael Spiteller, Manoj Ghaste, Rainer Wirth, Hubert Herz,
Dieter Spiteller and Jerrold Meinwald
Source: Proceedings of the National Academy of Sciences of the United States of America,
Vol. 108, No. 5 (February 1, 2011), pp. 1955-1960
Published by: National Academy of Sciences
Stable URL: http://www.jstor.org/stable/41001798
Accessed: 20-07-2018 12:03 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/41001798?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

National Academy of Sciences is collaborating with JSTOR to digitize, preserve and extend
access to Proceedings of the National Academy of Sciences of the United States of
America

This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
 Chemical basis of the synergism and antagonism
in microbial communities in the nests of
leaf-cutting ants
Ilka Schoenian3, Michael Spitellerb, Manoj Ghasteb, Rainer Wirthc, Hubert Herzd, and Dieter Spiteller3'1
aDepartment of Bioorganic Chemistry, Max Planck Institute for Chemical Ecology, D-07745 Jena, Germany; institute of Environmental Research of the Faculty
of Chemistry, Dortmund University of Technology, D-44221 Dortmund, Germany; department of Plant Ecology and Systematics, Technical University of
Kaiserslautern, D-67663 Kaiserslautern, Germany; and dSmithsonian Tropical Research Institute, Apartado 0843-03092, Balboa, Ancón, Republic of Panamá
Edited* by Jerrold Meinwald, Cornell University, Ithaca, NY, and approved November 22, 2010 (received for review June 16, 2010)
Leaf-cutting ants cultivate the fungus Leucoagaricus gongylo-
phorus, which serves as a major food source. This symbiosis is
threatened by microbial pathogens that can severely infect L
gongylophorus. Microbial symbionts of leaf-cutting ants, mainly
Pseudonocardia and Streptomyces, support the ants in defending
their fungus gardens against infections by supplying antimicrobial
and antif ungal compounds. The ecological rolé of microorganisms in
the nests of leaf-cutting ants can only be addressed in detail if their
secondary metabolites are known. Here, we use an approach for the
rapid identification of established bioactive compounds from micro-
organisms in ecological contexts by combining phylogenetic data,
database searches, and liquid chromatography electrospray ionisa-
tion high resolution mass spectrometry (LC-ESI-HR-MS) screening.
Antimycins Ai-A* valinomycins, and actinomycins were identified
in this manner from Streptomyces symbionts of leaf-cutting ants.
Matrix-assisted laser desorption ionization (MALDI) imaging
revealed the distribution of valinomycin directly on the integument
of Acromyrmex echinatior workers. Valinomycins and actinomycins
were also directly identified in samples from the waste of A. echi-
natior ana A. niger leaf -cutting ants, suggesting that the compounds
exert their antimicrobial and antifungal potential in the nests of
leaf-cutting ants. Strong synergistic effects of the secondary meta-
bolites produced by ant-associated Streptomyces were observed in
the agar diffusion assay against Escovopsis weberi. Actinomycins
strongly inhibit soil bacteria as well as other Streptomyces and Pseu-
donocardia symbionts. The antifungal antimycins are not only active
against pathogenic fungi but also the garden fungus L gongylopho-
rus itself. In conclusion, secondary metabolites of microbial sym-
bionts of leaf-cutting ants contribute to shaping the microbial
communities within the nests of leaf-cutting ants.
chemical imaging | antibiotics | chemical defense | ecological function
Leaf-cutting/fungus-growing
unique among ants, because they ants such
grow theasfungus
Acromyrmex
Leucoa- are
garicus gongylophorus (Agaricales: Leucocoprineae) with har-
vested leaf material in chambers of their nests (1). In turn,
L. gongylophorus is their major food source. However, this obli-
gate mutualistic interaction is threatened by various microbial
pathogens such as the fungi Escovopsis (2, 3), Fusarìum (A), Syn-
cephalastrum (4), and Trichoderma (4). In addition, microorganisms
from the surrounding soil or plant pathogens accidentally in-
troduced from harvested leaf material may compete with the
garden fungus for nutrients and living space (5). Therefore, leaf-
cutting ants treat their fungus gardens with great care, removing
any suspicious material into waste chambers (6). Besides this
mechanical cleaning behavior, leaf-cutting ants make use of an-
timicrobial chemicals (7-9); these include 3-hydroxydecanoic
acid, which is secreted from the ants' metapleural glands.
However, in 1999, Currie et al. (10) discovered microbial sym-
bionts, identified as Pseudonocardia, from biofilms on the in-
tegument of leaf-cutting ants. Because Pseudonocardia belong to
the well-known antibiotic-producing Actinobacteria, it was sus-
www.pnas.org/cgi/doi/10.1073/pnas.1008441108 PNAS | February 1, 2011 | vol.108 | no. 5 | 1955-1960
This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
pected that Pseudonocardia play a crucial role in the ants' defense
against pathogens. Although isolated microbial symbionts were
active against Escovopsis in the agar diffusion assay (10), until re-
cently, not a single compound from the ants' microbial symbionts
had been characterized. Using bioassay-guided isolation, Haeder
et al. (11) identified the antifungal candicidin macrolides that
are produced by a large number of Streptomyces symbionts isolated
from three different leaf-cutting ant species (A. octospinosus,
A. echinatior, and A. volcanus). For the fungus-growing ant Apter-
ostigma dentigerum, Oh et al. (12) reported the cyclodepsipeptide
dentigerumycin from a Pseudonocardia symbiont with activity
against E. weberi. From A octospinosus, Barke et al. (13) detected,
besides the presence of the previously characterized candicidin
polyene macrolide-producing Streptomyces (11), a Pseudono-
cardia symbiont that produces a nystatin-like polyene macrolide.
These recent findings indicated that there are likely to be a num-
ber of diverse antifungal compounds yet to be identified from
microbial symbionts of leaf-cutting/fungus-growing ants (11-13).
Over the last few years, researchers have realized that the eco-
system of leaf-cutting ants is much more complex than initially
described as a coevolution of the leaf-cutting ants, their fungus
garden L. gongylophorus, one microbial symbiont (Pseudono-
cardia), and one specialized fungal pathogen (Escovopsis) (14). In
addition to the characterization of microbial antifungal compounds
(11-13) involved in mediating the interactions between the differ-
ent partners, it has become evident that both a large number of
pathogens (2-5) pose a threat to the ants' fungus garden and
a large diversity of microbial symbionts can be found within the
ants' nests. These symbionts fulfill diverse functions, including de-
fense against pathogens or promotion of the growth of the garden
fungus (e.g., by nitrogen fixation) (10-13, 15-17). Because some
bacterial symbionts of leaf-cutting ants can be also detrimental to
the growth of the mutualistic fungus L. gongylophorus, the one-
sided view of leaf -cutting ants and Actinomyces symbionts as mu-
tualistic partners of the leaf-cutting ants should be challenged (18).
In order to begin to better understand the ecological role of
microorganisms associated with leaf-cutting ants, it is crucial to
expose the chemistry of the individual (micro)organisms in the
community. Extending phylogenetic comparisons of secondary
metabolite producers (19-21), we used a combination of phylo-
genetic analysis, database screening, and electrospray ionisation
high resolution mass spectrometry (ESI-HR-MS) analysis to
UJ
s
Author contributions: I.S. and D.S. designed research; I.S., M.S., M.G., R.W., H.H., and D.S.
performed research; I.S. and D.S. analyzed data; and I.S. and D.S. wrote the paper.
The authors declare no conflict of interest.
*This Direct Submission article had a prearranged editor.
Data deposition: The sequences reported in this paper have been deposited in the Gen-
Bank database (accession nos. HM538453 and HM538454). O
I
1To whom correspondence should be addressed. E-mail: dspiteller@ice.mpg.de.
u
UJ
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1 073/pnas.1 008441 108/-/DCSupplemental.
rapidly identity ecologically relevant secondary metabolites from
microbial symbionts. In this way, we identified several antibiotics
and investigated their role in shaping the complex interactions in
the leaf-cutting ants' ecosystem.
Results
Structure Elucidation of Antimicrobial and Antifungal Compounds
from Microbial Symbionts. A variety of microbial symbionts from
three different Acromyrmex leaf-cutting ant species (A. octospi-
nosus, A. echinatior, and A. volcanus) had been isolated in a pre-
vious study. The symbiotic microorganisms were characterized
using their 165 rDNA sequences for database comparison (11).
In addition, many more 16S rDNA sequences from microbial
symbionts of leaf-cutting ants are now available (11, 16, 22-24).
However, because most research concerning the symbionts of
leaf-cutting ants has focused on Pseudonocardia, their 165 rDNA
sequences are the main ones that have been collected and used to
study the evolution of the symbiosis between Pseudonocardia and
leaf-cutting ants (14, 23).
Instead of using the 165 rDNA sequences for evolutionary
studies (19-21), we used phylogenetic data as a guideline to
rapidly identify secondary metabolites that might play a crucial
role in the interactions of the complex microbial community of
leaf-cutting ants. We identified the closest well-studied relatives
to Streptomyces symbionts from leaf-cutting ants based on their
165 rDNA sequence similarity to sequences from the Greengenes
and National Center for Biotechnology Information (NCBI)
databases using the blast algorithm. The secondary metabolite
production of the identified relatives was then studied using the
Chemical Abstracts Service (CAS) SciFinder database. On the
basis of the results from this phylogenetic comparison, culture
supernatants and methanol extracts of the microbial symbionts
from Acromyrmex ants were screened by liquid chromatgraphy
mass spectrometry (LC-MS) in a search for the [M+H]+ ions of
suspected natural products. Putative hits were verified by com-
paring the retention times and ESI-HR-MS spectra of selected
compounds with those of authentic standards.
For example, Streptomyces sp. Av25_2 showed high similarity
to Streptomyces parvus str. NBRC 14599 (AB184603.1; 99.57%)
(Table SI and Fig. SI). Another 5. parvus strain had previously
been characterized as a producer of actinomycin C (25). In light of
this information, we used LC-MS to screen Streptomyces sp.
Av25_2 for actinomycin production. Indeed, the symbiotic strain
Streptomyces sp. Av25_2 was found to produce actinomycin D (1)
and the closely related actinomycin X2 (2) (26, 27) (Fig. 1, Table SI,
iitVi NH2 Vo 0:s^
I I / HN-f V^ '
Actinomycin D:R = CH21 >- O 0 J,H »
Actinomycin X2: R = C=O 2 - -Ï j o '~J
O f 3V"( )-aR
i >Y» i I 2 Valinomycin
O S3). In our bioassays, valinomycins inhibited only the growth of
Antimycin A^ R^ -C6H13; R2 -CO-C4H9 4
Antimycin A2: R^ -C6H13; R2 -CO-C3H7 5
Antimycin A3: R, -C4H9; R2 -CO-C4H9 6
Antimycin A4: F^ -C4H9; R2 -CO-C3H7 7
Fig. 1. Structures of actinomycin D (1), actinomycin X2 (2), valinomycin (3),
and antimycins Ai-^ (4-7) produced by Streptomyces associated with leaf-
cutting ants.
1 956 | www.pnas.org/cgi/doi/i 0. 1 073/pnas. 1 008441 1 08
This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
and Fig. SI). The identity of actinomycin D (1) and X2 (2) was
further confirmed by NMR and comparison with commercially
available actinomycin standards. Valinomycin (3) (28) and its
closely related derivatives (29) were identified from the symbiont
Streptomyces sp. Av25_3, revealing its high similarity to valino-
mycin producers (Fig. 1, Table SI, and Fig. SI) such as S. anulatus
(EU647474.1; 98.84%) (Table SI and Fig. SI) and S. tsusimaensis
(EU622279; 99.21%) (Table SI and Fig. SI) (20). Furthermore,
we identified the well-known antifungal antimycins A!-A4 (4-7)
(30) (Fig. 1, Table SI, and Fig. SI), which are produced by mi-
crobial symbionts of Acromyrmex, on the basis of the high simi-
larity of Streptomyces sp. AolO to S. albidoflavus (DQ855477.1;
99.72%) (Table SI and Fig. SI). S. albidoflavus, which was re-
cently isolated from mangroves, has been shown to produce
antimycin A18 (31), suggesting the possible production of anti-
mycins (4-7) by Streptomyces AolO.
The LC-MS screening of the extracts from all Streptomyces
isolates (11) from leaf-cutting ants revealed the distribution of
the identified antibiotics in the microbial isolates from three
Acromyrmex species (A. octospinosus, A. volcanus, A. echinatior).
Valinomycins were produced by Streptomyces sp. Av25_3, Strep-
tomyces sp. Av26_3, and Streptomyces sp. Av25_6. Antimycins
seem to be widespread among Streptomyces and were found in one-
half of the Streptomyces symbionts analyzed, whereas actinomycins
were only produced by Streptomyces sp. Av25_2 (Table SI).
Function of Secondary Metabolites from the Symbionts of Leaf-
Cutting Ants. To evaluate the ecological role of the identified sec-
ondary metabolites from the microbial symbionts of Acromyrmex
leaf-cutting ants, we tested the compounds in agar diffusion assays
against fungal pathogens of the fungus garden (E. weberi, F.
decemcellulare, and T. harzianum) and insect pathogenic fungi
(Metarhizium anisopliae, Beauveria bassiana, and Cordyceps mili-
taris). The black yeast Phialophora fastigiata (32) and the fungus
Syncephalastrum racemosum (4), both of which are known patho-
gens in the nests of leaf-cutting ants, were included in the agar
diffusion assays as well. We also examined the inhibitory potential
of actinomycins (1, 2), valinomycin (3), valinomycin derivatives,
and antimycins (4-7) against the common soil bacterium Bacillus
subtilis. In addition, the effects of the bioactive substances identi-
fied (1-7) were also investigated (Fig. 1) on selected nonproducing
microbial isolates from Acromyrmex leaf-cutting ants and the mu-
tualistic fungus L. gongylophorus. The results of the bioassays are
presented in Table 1, Table S2, and Figs. S2-S8.
Actinomycin D (1) strongly inhibited B. subtilis (Fig. S5) as well
as the growth of Streptomyces sp. Av25_4 (Fig. S6) but not
Streptomyces sp. AolO associated with the leaf-cutting ants (Fig.
S6). Even as little as 0.4 nmol actinomycins (1, 2) hampered the
growth oí Pseudonocardia sp. Aol and Pseudonocardia sp. Av30
(Fig. S6) in the agar diffusion assay. Greater quantities of acti-
nomycins also inhibited the growth of the fungi F. decemcellulare,
S. racemosum, and the black yeast P. fastigiata (Fig. S4).
The well-known antifungal antimycins (58 nmol) (4-7) created
inhibition zones in the agar diffusion assay against E. weberi (2,
3) (Fig. S2), the black yeast P. fastigiata (32) (Fig. S4), and
G militaris but not against other insect pathogenic fungi (e.g.,
B. bassiana) tested.
As little as 5.8 nmol antimycins A1-A4 (4-7) clearly inhibited
the growth of the leaf-cutting ants' mutualistic fungus L. gongy-
3
lophorus in the agar diffusion assay. In contrast, valinomycins (up
to 240 nmol) did not inhibit the growth of L. gongylophorus (Fig.
B. subtilis (Fig. S5).
Mixtures of actinomycins (1, 2), valinomycin (3), valinomycin
derivatives, antimycins (4-7), and candicidins (11) exhibit strong
synergistic effects (33, 34). Combined with valinomycins or anti-
mycins (4-7), candicidin macrolides close to or below the minimal
inhibitory concentration (MIC, around 1-6 nmol each) caused
Schoenian et al.
Table 1. Inhibition zones caused by antimycins Ai-A4 (4-7), actinomycin D (1) and X2 (2), and valinomycin (3) and valinomycin
derivatives in the agar diffusion assay against fungi and bacteria
Antimycins (nmol/10 rnL Actinomycins (nmol/10 mL Valinomycins (nmol/10 mL
Organism SFM cm inhibition zone) SFM cm inhibition zone) SFM cm inhibition zone)
Leucoagaricus gongylophorus 5.8/58 2.4/24 2.7/27
1.0/1.7 X/0.2 X/X
Escovopsis weberi 5.8/58 2.4/24 2.7/27
0.3/0.4 X/X X/X
Phialophora fastigiata 58/280 1.2/240 2.7/270
0.1/0.2 X/0.2 X/X
Trichoderma harzianum 58 24 27
0.3 X X
Beauveria bassiana 27.5 4/24 1.4/270
X X/X X/X
Metarhizium anisopliae 27.5/58 4/24 1.4/270
X/X X/X X/X

Cordyceps militaris 27.5/58 4/24 1.4/270

0.1/0.2 X/X X/X
Fusarium decemcellulare 580 240 270
X 0.1 X

Streptomyces sp. Av 25_4 27.5 4/12 1.4

X 0.2/0.8 X

Streptomyces sp. Av 2 5_2 27.5 12 -

XX -

Streptomyces sp. Ao 10 27.5 4 1.4

XXX

Pseudonocardia sp. Ao1 27.5 4/12 1.4

X 0.1/0.8 X

Pseudonocardia sp. Av30 27.5 0.4/12 1.4/2.7

X 0.1/0.6 X/X
Bacillus subtilis 275 0.4/2/12 270
X 0.3/0.8/1.1 0.2

Syncephalastrum racemosum 145 120 120

X 0,1 X

SFM, soy flou

clear on the body, particularly, the alitrunk, but also on the legs.
inhibit
(Table
Judging from the S2MALDI imaging, A. echinatior anants can have
several nanograms of valinomycin (3) on their cuticle. It is note-
Directworthy thatScreen
valinomycin (3) seems to be often produced in highest
in the Waste
amounts in highly localized patches (Fig. 2).
directly asse
tifiedDiscussion from
LC-MS to
Phylogenetic analysis is widely used to classify ide
microorganisms
material. Fu
and more recently, to study evolutionary relationships among
waste mater
secondary metabolite producers (19-21). Here, we show that 16S
A. echinatio rDNA data combined with database search and LC-MS profiling
ethyl acetat
can be highly valuable for quickly identifying secondary meta-
methanol, bolite profiles from new microbial isolates. This straightforward a
was detected
method has great potential to help reveal the chemistry in vari-
contrast, ous complex microbial associations. va
ternal calibration curve was used to estimate the amounts of Using the '6S rDNA analysis combined with database searching
and LC-MS screening, we have identified valinomycins, actino-
valinomycins and actinomycins present in the waste of leaf -cutting
ants. Actinomycins were found at concentrations of the order mycinof D (1), actinomycin X2 (2), and antimycins Ai-A* (4-7) (Fig.
170 pmol/g. The quantity of valinomycins in the various waste1, Table SI, and Fig. SI) as compounds that are produced by mi-i
samples varied considerably, ranging from 0 to 13 nmol/g.crobial
Acti- symbionts of Acromyrmex ants. Actinomycins (1, 2) were
u

nomycins (1, 2) and valinomycins were not detected in fungus found only from a single A. volcanus symbiont Streptomyces sp.
garden samples. In the limited sample material available to Av25_2,
us, we whereas valinomycins and antimycins A1-A4 (4-7) were
produced
were unable to directly detect antimycins A1-A4 (4-7) in waste or by several of the Streptomyces symbionts analyzed.
fungus garden samples. Similarly, candicidin macrolides were synthesized by the majority
Valinomycin (3) was detected directly on the integument of Streptomyces isolates from the three Acromyrmex species >-
of the
o
A. echinatior workers using matrix-assisted laser desorption ioni- (11). The broad occurrence of those antibiotics suggests
tested
zation (MALDI) imaging. Valinomycin (3) was found in varying their important role in the ecosystem of the Acromyrmex ants.
concentrations and at various positions on the ants' bodies.The Forantifungal antimycins Ax-A4 (4-7) inhibit the electron
transfer of the ubiquinolxytochrome c reducíase (complex III)
some A. echinatior workers, valinomycin (3) was not only detected

Schoenian et al. PNAS | February 1, 2011 | vol.108 | no. 5 | 1957

This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
 SpESÌBS^r ÌRiS9^B[ HHSSs^H
2^2&S^S ^^B^^H nSSEsm
^S^^H^^B S^BS^EmI SfiSSS^^Si
35^33ES£ SSDBSSsi SS^SI
Ö i Y (micron) Vinieron) Y (micron)
0 1000 2000 3000 4000 5000 6000 7000 8000 900010000 o _iooo 2000 aooo «ooo sooo

fi^^S^l fìl^B
■■n l^^HH !^K^9 1
i^^^^^^^H^BI i^^^^^^^HBBf i^^m^l^^^H Imin
Fig. 2. (A) Microscopic pictures of A. echinatior workers mounted onto MALDI plates. Beside the ants are valinomycin spots as reference to estimate the
amount of valinomyin (3) on the integument of leaf-cutting ants (red circles). (Scale bars in the microscope images, 1 mm.) (B) MALDI images reflect the
distribution of valinomycin (3) ([M+K]+ m/z = 1,149) on the integument of the A. echinatior workers shown in A. The color code of the heat maps is logarithmic
and corresponds to different concentrations of valinomycin (3). During the MALDI analysis, the abdomen of all ants fell off, and therefore, there are no data
of the valinomycin distribution on the abdomen available.

of the respiratory chain and thus induce apoptosis (30, 35). Thus,
antimycins A1-A4 (4-7) inhibit Candida albicans, Mucor mucedo,
and Aspergillus niger; however, they are not particularly active
against Fusarium (36), which is a pathogen in the ants' nests. In
addition, antimycins (4-7) were found to inhibit the growth of
several bacteria (37) and yeasts (35). Antimycins A1-A4 (4-7)
inhibited the growth of E. weberi (Table 1) and the detrimental
black yeast P. fastidiata (32) in our agar diffusion assays against
pathogens of the leaf-cutting ants. In addition, we observed
a strongly antagonistic effect of antimycins A1-A4 (4-7) against
the mutualistic fungus L. gongylophorus . This provides a chemical
explanation for recent observations, namely that the microbial
symbionts of leaf-cutting ants have negative as well as positive
effects on the ant colony (18, 23). In the case of fungal infection,
the cost of inhibiting growth of the garden fungus L. gongylo-
phorus may be outweighed by the benefit of preventing the
spread of infection in the nest.
Valinomycin (3) (28) and structural variants (29) are ionophores
that selectively bind potassium ions, leading to the disintegration of
the cellular membrane potential and thus, destroying cells (38).
However, in the agar diffusion assays, valinomycins did not inhibit
the fungi tested, including the symbiotic garden fungus (Table 1).
Nevertheless, valinomycin (3) is known to inhibit the hyphal growth
of C. albicans (39) as well as various phytopathogens (40, 41). The
ability to defend against phytopathogens is potentially relevant,
because leaf-cutting ants may import these pathogens with the leaf
material that they collect to feedL. gongylophorus (5). Additionally,
valinomycin (3) is active against insects, nematodes, and mites (42).
Actinomycins have been identified as highly active antibiotics
(43), but on account of their high toxicity, actinomycins are used
exclusively in cancer therapy (44). Because actinomycins strongly
inhibit the growth of competing microorganisms, including other
Streptomyces and Pseudonocardia symbionts of leaf-cutting ants
(Table SI), they obviously help Streptomyces sp. Av25_2 compete
with neighboring microorganisms. Thus, with actinomycins iden-
tified as being synthesized by a Streptomyces strain isolated from
leaf-cutting ants, we provide the chemical basis for the observation
of Sen et al. (18), namely that ant- associated microorganisms play
different roles in this ecosystem, ranging from mutualistic to det-
rimental interactions. Competition between their different mi-
crobial symbionts may, at first glance, be disadvantageous for the
leaf-cutting ants; nevertheless, it may provide a pathway for the
selection of potent defenders by inducing an evolutionary arms
race between them.
Because antimycins A1-A4 (4-7), actinomycins, and valinomy-
cins as well as the previously identified candicidin macrolides (11)
are likely to occur in concert in the leaf-cutting ants' nest, we in-
vestigated the inhibitory properties, of different compound mix-
tures. We found, indeed, that as mixtures, these antibiotics inhib-
ited E. weberi growth more efficiently than the single compounds.
Amounts close to or below the minimal inhibitory concentration of
each compound together caused clear inhibition zones in the agar
diffusion assay (Figs. S7 and S8). Such synergistic effects between
valinomycin (3) or actinomycin D (1) with polyene macrolides have
been observed previously in pharmacological studies (33, 34). Now,
we show that chemical diversity, particularly, the interplay of bio-
active molecules, is an important factor in the protection of the leaf-
cutting ants' nests against infections.
Using MALDI imaging, we were able to identify and monitor
the local distribution of an antibiotic, valinomycin (3), from mi-
crobial symbionts of leaf-cutting ants directly on the integument
(Fig. 2). On the cuticle of A. echinatior workers, patches with high
amounts of valinomycin (3) (e.g., at joints of the legs) were
detected. Considering the weight of the A. echinatior workers used
for the MALDI imaging (~6 mg), several nanograms of valino-
mycin (3) and in some patches, several tenths of nanograms are
likely to be sufficient to fight against susceptible organisms. Until
now, it has only been possible to directly detect antibiotics of mi-
crobial symbionts from insects on the cocoon of beewolf larvae
(45). The presence of valinomycin (3) on the ants' bodies suggests
that it may play an important role in protecting individual workers,
probably not only against microbial pathogens but also against
parasites (e.g., mites) (1, 42). In addition, we found actinomycins
(1, 2) and valinomycin (3) in the waste oL4. niger and A echinatior,
which shows again that Streptomyces sp. play an essential role (11,
16) in the environment of leaf-cutting ants. The high variability in
the quantities of valinomycin (3) in the waste oí A. niger provides
molecular proof that the antibiotics of the microbial symbionts of
leaf-cutting ants can be highly localized (46). Therefore, it seems
possible that the production of antibiotics may be regulated by the
producing microorganisms or even the leaf-cutting ants (46). The
direct detection of antibiotics in natural environments is often
problematic, because these compounds can be active in very low
concentrations and their occurrence can vary both locally and

1958 | www.pnas.org/cgi/doi/10.1073/pnas.1008441108 Schoenian et al.

This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
temporally (47). The limited quantities of fungus garden and waste Actinomycins (1, 2) from Streptomyces sp. Av25_2. Based on the 16S rDNA
samples available to us may explain why we found only valinomy- similarity of Streptomyces sp. Av25_2 to S. pan/us (99.57%), which is a
known producer of actinomycin C (25), a methanol extract from a Strep-
cins and in some cases, actinomycins (1, 2) in our extracts.
tomyces sp. Av25_2 culture (6 mL) was screened for actinomycin production
Leaf-cutting ants likely benefit from the rich diversity of anti-
by LC-MS. Retention time of actinomycin D (1): 24.8 min (HR-ESI-MS: [M+H]+
microbial and antifungal secondary metabolites from their mi- measured: 1,255.6353, calculated: 1,255.6363 C62H87Ni2Oi6); retention time
crobial symbionts, because combined, these antibiotics can have of actinomycin X2 (2): 25.2 min (ESI-HR-MS: [M+H]+ measured: 1,269.6143,
strong synergistic effects against possible threats; however, these calculated: 1,269.6156 Q^N^O^).
compounds can also have detrimental side effects. The diversity of Methanol extracts of all microbial symbionts (1 1) from Acromyrmex leaf-
natural products plays a driving role in shaping the ecosystems cutting ants were screened for actinomycin production using LC-MS.
of leaf-cutting ants. Only by revealing their chemical nature we
can begin to understand fully the complex interactions between Valinomycin (3) and Valinomycin Derivatives from Streptomyces Symbionts of
Acromyrmex. Streptomyces sp. Av25_3 exhibited high similarity to known
multiorganismic partners. The combined approach of phyloge-
valinomycin producers S. tsusimaensis (99.21%) and 5. anulatus (98.84%).
netic analysis with chemical analytics presented here can signifi-
Methanol extracts of Streptomyces sp. Av25_3 were screened by LC-MS and
cantly speed up the identification of the diverse bioactive natural compared with the retention time of an authentic standard of valinomycin.
products that orchestrate the multiple interactions in complex Retention times of valinomycin (3) and valinomycin derivatives (29):
biological systems. valinomycin-28 amu: 41.5 min (ESI-HR-MS: [M+NH4]+ measured: 1,100.6338,
calculated: 1,100.6342 C52H9oN7018; ESI-HR-MS: [M+K]+ measured: 1,121.5621,
Materials and Methods
calculated: 1,121.5630 C52H86N6O18K); valinomycin-14 amu: 43.6 min (ESI-HR-
Fungal and Microbial Cultures. E. weberi CBS 1 10660 was obtained fromMS: the[M+NH4]+ measured: 1,1 14.6499, calculated: 1,1 14.6493 CsaHgzNyO^; ESI-
Centralbureau voor Schimmelcultures in Utrecht, The Netherlands. L gon- HR-MS: [M+K]+ measured: 1,135.5776, calculated: 1,135.5787 C53H88N6O18K);
gylophorus was an isolate from the fungus garden of Atta colombica (lab-valinomycin (3): 46.3 min (ESI-HR-MS: [M+NH4]+ measured: 1,128.6658, cal-
oratory colony collected in Gamboa, Panama, in 2004). F. decemcellulare culated:
was 1,128.6650 C54H94N7O18; ESI-HR-MS: [M+K]+ measured: 1,149.5936,
from the Phytopathology Department of the Friedrich Schiller University calculated: 1,149.5949 C54H9oN6018K); valinomycin+14 amu: 49.7 min (ESI-HR-
Jena, and B. bassiana FSU 5084 was obtained from the PilzreferenzzentrumMS: [M+NH4]+ measured: 1,142.6808, calculated: 1,142.6806 C55H94N7O18; ESI-
of the Friedrich Schiller University Jena (Jena, Germany). T. harzianum HR-MS:
DSM [M+K]+ measured: 1,163.6087, calculated: 1,163.6100 CssHgoNeO^K).
63059, P. fastigiata DSM 2692, M. anisopliae DSM 1490, C militaris DSMMethanol extracts of all microbial symbionts (11) from Acromyrmex leaf-
1153, B. subtilis DSM 10, and S. racemosum DSM 859 originated fromCutting
the ants were screened for valinomycin production using LC-MS.
German Collection of Microorganisms and Cell Cultures (Braunschweig,
Germany). The Streptomyces and Pseudonocardia strains used for bioassays Antimycins At-A4 (4-7) from Streptomyces Symbionts of Acromyrmex. 5. albi-
were isolates from A. volcanus, A. octospinosus, and A. echinatior leaf-cutting
doflavus, a producer of the antifungal antimycins such as antimycin A18 (31),
ants (11). All strains were maintained on soy flour medium (SFM) agar plates
exhibited high similarity to several Streptomyces symbionts from the leaf-
(20 g soy flour, 20 g mannitol, 15 g agar, 1 L ddH2O) (48). cutting ants: Streptomyces sp. Ae32j2 (5. albidoflavus, AJ002090.1;
100.00%), Streptomyces sp. Ao10 (S. albidoflavus, DQ855477.1; 99.72%),
Leaf-Cutting Ants and Fungus Garden Samples. Leaf -cutting ants and fungus
Streptomyces sp. Av28_2 (S. albidoflavus str. UST0407 11-291, FJ591 130.1;
garden samples from A. volcanus, A. octospinosus, and A. echinatior colonies
98.23%), Streptomyces sp. Av28_3 (S. albidoflavus, AJ002090.1; 99.07%),
were collected and identified in 2007 in Gamboa (Panama) by H.H. In addition,
Streptomyces sp. Av25_1 (S. albidoflavus, AJ002090.1; 97.78%), and Strep-
microorganisms were isolated from the fungus garden of an established tomyces
lab- sp. Av26_5 (S. albidoflavus, AJ002090.1; 99.93%).
oratory colony of A echinatior collected in 2002 by H.H. in Panama. Fungus Retention time of antimycin A^ (4): 29.0 min (ESI-HR-MS: [M+H]+ mea-
garden samples and waste samples were collected from A. echinatior sured:
and 549.2805, calculated: 549.2812 C28H41O9N2); retention time antimycin
A. niger laboratory colonies (collected in Brazil in 1999 and maintained
A2as(5): 28.0 min (ESI-HR-MS: [M+H]+ measured: 535.2651, calculated:
laboratory colonies by R.W.). 535.2656 C27H39O9N2); retention time antimycin A3 (6): 27.0 min (ESI-HR-MS:
[M+H]+ measured: 521.2493, calculated: 521.2500 C26H37O9N2); retention
time antimycin A4 (7): 26.0 min (ESI-HR-MS: [M+H]+ measured: 507.2342,
Cultivation of Microorganisms. Streptomyces from Acromyrmex leaf-cutting
calculated: 507.2343 C25H35O9N2).
ants were isolated and determined by Haeder et al. (11). All Streptomyces
Methanol extracts of all microbial symbionts (11) from Acromyrmex
were maintained on SFM agar plates (48). To obtain methanol extracts, the
microorganisms were grown in 20 mL test tubes fitted with springs for leaf-cutting
aer- ants were screened for antimycin Aì-A4 (4-7) production
ation containing 6 mL liquid SFM. The cultures were incubated at 28 °C on an LC-MS.
using
orbital shaker (220 rpm, Infors Multitron II MT25) for 6 d; 2 mL culture were
lyophilized and redissolved in 1 mL methanol. For larger-scale isolation of
Antimicrobial and Antifungal Properties of Antimycins A!-A4 (4-7),
secondary metabolites, 200 mL liquid SFM were filled into 500-mL Erlenmeyer
Actinomycins (1, 2), Valinomycin (3), and Valinomycin Derivatives. E. weberi,
flasks fitted with springs for aeration. The flasks were inoculated with
L. gongylophorus, F. decemcellulare, B. bassiana, T. harzianum, P. fastigiata, M.
a Streptomyces spore suspension. The cultures were grown in an orbital
anisopliae, C. militaris, S. racemosum, B. subtilis, Streptomyces sp. Av25_2,
shaker (220 rpm at 28 °C, Infors Multitron II MT25) for 6 d. Streptomyces sp. Av25_4, Streptomyces sp. Ao10, Pseudonocardia sp. Ao1, and
Pseudonocardia sp. Av30 were used as test organisms in the agar diffusion
Secondary Metabolite Screening Inspired by Phylogenetic Data. 165 rDNA assay
data against antimycins A1-A4 (4-7) (5.8-580.0 nmol), actinomycins (1, 2) (0.4-
240.0 nmol), and valinomycins (1.4-270.0 nmol). In addition, the performance
analysis was performed for Streptomyces symbionts obtained previously from
three Acromyrmex leaf-cutting ant species (11). Highly similar sequencesofus-E. weberi in the presence of compound mixtures was investigated.
ing blast algorithm were identified using the Greengenes (http://greengenes. For the bioassays, 100 'il mycelium or spore suspensions (~5 mg wet
Ibl.gov/cgi-bin/nph-index.cgi) (49) and NCBI (http://www.ncbi.nlm.nih.gov/
weight/mL in water) of the test organisms were spread onto SFM plates (5.5 >
cm diameter, 10 mL medium). A 6-mm hole was cut in the middle of the
nuccore) databases. The hits for well-studied, closely-related microorganisms
with similarities higher than >98.5% were screened for their antimicrobialplate to apply 50 'il test solutions or an appropriate solvent control (MeOH).
and antifungal secondary metabolites using SciFinder from the CAS.The Theinhibition zones were monitored after samples were incubated for 4-16 d
construction of phylogenetic trees was performed with the program MEGA at 28 °C All assays that showed inhibition zones were performed at least in
version 4 using the neighbor-joining method (bootstrap value, n = 1,000) triplicate
(50). and compared with identically prepared solvent controls. The results
of the assays are presented in Table 1, Table S2, and Figs. S2-S8.
LC-ESI-MS Analysis of Extracts from Microbial Symbionts. Fifteen microliters of
extract were injected into a Dionex Ultimate 3000 HPLC system fitted with
Detection of Antibiotic and Antifungal Compounds in the Fungus Garden and
a Phenomenex Kinetex C18 column (2.6 ^m, 100 Á, 150 x 2.1 mm). Either Waste. Waste and fungus garden samples from A echinatior and A. niger
a Thermo Fisher LTQ or an LTQ Orbitrap with an ESI ion source served as an analyzed. The samples (0.58-31 .81 g) were extracted with ethyl acetate
were
MS detector. HPLC conditions: 3 min in 100% A, 27 min to 100% B, 10 min in mL) by sonif ¡cation and stirring for 2 h. After filtration, the filtrate
(10-300
100% B (A, H2O 0.5% AcOH; B, MeCN 0.5% AcOH; flow rate = 0.20 mL/min).was concentrated in vacuo. The residue was resuspended in 200 'il, 500 |iL,
Compounds were identified by their retention times and HR-ESI-MS/MS or 2,500 [xL methanol and analyzed by LC-MS. The measurements were
spectra compared with commercially available standards. compared with standards of actinomycin D (1) and X2 (2), antimycins A^A4

Schoenian et al. PNAS | February 1, 2011 | vol.108 | no. 5 | 1959

This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms
(4-7), valinomycin (3), and valinomycin derivatives. The total amounts of
valinomycins and actinomycins in the extracts were estimated by comparison
of the respective peak areas to an external calibration curve.
MALDI Imaging of A echinatior Workers. A. echinatior workers were mounted
onto the MALDI plate with the help of a double-sided adhesive tape and
sprayed with matrix solution of oc-cyano^-hydroxycinnamic acid (51) [7 mg/mL
in MeCN:water (80:20, v:v) and 0.2% trifluoroacetic acid] by an automatic
sprayer (Bruker ImagePrep). Microscopic pictures (magnification = 1-1 .25x)
were taken using a Leica S8 APO Greenough stereo microscope equipped with
a Schott KL 1 500 compact halogen cold light source. The images were captured
by a digital camera and were processed using the Leica Application Suite LAS
EZ ver. 1.6.O. All ¡mages were rescaled to the same scale ratio to make parity
among all of the different images. MALDI imaging experiments of antibiotics
on the cuticle of A. echinatior workers were performed by an LTQ-Orbitrap XL
mass spectrometer (Thermo Fisher) coupled to a MALDI source to provide
spectra and images. The spectrometer was operated in positive-selected ion
monitoring mode (mass range = 1,060-1,160) with nominal mass resolving
power of 60,000 at m/z 400 and a scan rate of 1 Hz with automatic gain control
to provide high-accuracy mass measurements within 2-ppm deviation using
the internal calibration standard a-cyano^-hydroxycinnamic acid (CHCA):m/z =
379.095. The laser was set at power 20 'i), with raster plate motion and raster
step size of 1 00 |im; three microscans per step are used for the analysis. All data
were processed by Thermo ImageQuest 1.0.1 software.
To estimate the amount of valinomycin (3) ([M+K]+ m/z = 1,149.602) on
the integument of leaf-cutting ants, a dilution series of valinomycin (3) (1,
10, and 100 ng) was measured under identical conditions as the ant samples.
ACKNOWLEDGMENTS. We thank the Smithsonian Tropical Research In-
stitute (STRI) for providing logistic help and the Autoridad Nacional del
Ambiente y el Mar (ANAM) for permission to sample ant colonies in Panama
and for issuing export permits. We are indebted to Dr. Kusari for performing
the stereomicroscope imaging and Emily Wheeler for editorial assistance.
D.S. thanks Professor Dr. Boland for his generous support, the Deutsche
Forschungsgemeinschaft for an Emmy Noether Fellowship (SP 1106/3-1),
and the Jena School for Microbial Communication (JSMC) for financial sup-
port. Further funding by the Verband der Chemischen Industrie and the Max
Planck Society is gratefully acknowledged.

1. Hölldobler B, Wilson EO (1990) The Ants (Springer, Berlin). 26. Bullock E, Johnson AW (1957) Actinomycin 5. The structure of actinomycin D. J Chem
2. Seifert KA, Samson RA, Chapela IH (1995) Escovopsis aspergilloides, a rediscovered Soc 3280-3285.

hyphomycete from leaf-cutting ant nests. Mycologia 87:407-413. 27. Brockmann H, Manegold JH (1960) Actinomycine. 23. Antibiotica aus Actinomycete
3. Currie CR, Mueller UG, Malloch D (1999) The agricultural pathology of ant fungus 45. Überführung von Actinomycin X2 in die Actinomycine C1f X0-ß und X0-ô. Chem Be
gardens. Proc Nati Acad Sci USA 96:7998-8002. 93:2971-2982.

4. Rodrigues A, et al. (2005) Variability of non-mutualistic filamentous fungi associated 28. Shemyakin MM, Aldanova NA, Vinogradova El, Feigina MY (1963) The structure a
with Atta sexdens rubropilosa nests. Folia Microbiol (Praha) 50:421-425. total synthesis of valinomycin. Tetrahedron Lett 1921-1925.
5. Fisher PJ, Stradling DJ, Sutton BC, Petrini LE (1996) Microfungi in the fungus gardens 29. Pitchayawasin-Thapphasaraphong S, Isobe M (2006) Molecular diversity of valinomyc
of the leaf-cutting ant Atta cephalotes: A preliminary study. Mycol Res 100:541-549. a 36-membered cyclic depsipeptide, which was detected by means of HPLC-Q-TOF-M
6. Bot ANM, Currie CR, Hart AG, Boomsma J (2001) Waste management in leaf-cutting on hydrolysate dipeptides. ITE Letters on Batteries. New Technol Med 7:465^473.
ants. Ethol Ecol Evol 13:225-237. 30. van Tamelen E, Strong FM, Loomans ME, Dickie JP, Dewey RS (1961) Chemistry o
7. Schildknecht H, Koob K (1971) Myrmicacin, the first insect herbicide. Angew Chem Int antimycin A. 10. Structure of antimycins. J Am Chem Soc 83:1639-1646.
Ed Engl 10:124-125. 31. Yan LL, et al. (2010) Antimycin A18 produced by an endophytic Streptomyce
albidoflavus isolated from a mangrove plant. J Antibiot (Tokyo) 63:259-261.
8. Bot ANM, Ortius-Lechner D, Finster K, Maile R, Boomsma JJ (2002) Variable sensitivity
of fungi and bacteria to compounds produced by the metapleural glands of 32.leaf-
Little AEF, Currie CR (2007) Symbiotic complexity: Discovery of a fifth symbiont in th
cutting ants. Insectes Soc 49:363-370. attine ant-microbe symbiosis. Biol Lett 3:501-504.
9. Fernandez-Marin H, Zimmerman JK, Rehner SA, Wcislo WT (2006) Active use33.
of Kwan
the CN, Medoff G, Kobayashi GS, Schlessinger D, Raskas HJ (1972) Potentiation of th
metapleural glands by ants in controlling fungal infection. Proc R Soc Lond B Biolantifungal
Sci effects of antibiotics by amphotericin B. Antimicrob Agents Chemother 2:
61-65.
273:1689-1695.

10. Currie CR, Scott JA, Summerbell RC, Malloch D (1999) Fungus-growing 34. Bolard J (1986)
ants use How do the polyene macrolide antibiotics affect the cellular
membrane properties? Biochim Biophys Acta 864:257-304.
antibiotic-producing bacteria to control garden parasites. Nature 398:701-704.
35. Kluepfel D, Sehgal SN, Vézina C (1970) Antimycin A components. I. Isolation and
11. Haeder S, Wirth R, Herz H, Spiteller D (2009) Candicidin-producing Streptomyces
biological activity. J Antibiot (Tokyo) 23:75-80.
support leaf-cutting ants to protect their fungus garden against the pathogenic
36. Ueki M, et al. (1997) UK-3A, a novel antifungal antibiotic from Streptomyces sp. 517-02:
fungus Escovopsis. Proc Nati Acad Sci USA 106:4742-4746.
Fermentation, isolation, structural elucidation and biological properties. J Antibiot
12. Oh DC, Poulsen M, Currie CR, Clardy J (2009) Dentigerumycin: A bacterial mediator of
(Tokyo) 50:551-555.
an ant-fungus symbiosis. Nat Chem Biol 5:391-393.
37. Marquis RE (1965) Nature of bactericidal action of antimycin A for Bacillus
13. Barke J, et al. (2010) A mixed community of actinomycetes produce multiple
megaterium. J Bacteriol 89:1453-1459.
antibiotics for the fungus farming ant Acromyrmex octospinosus. BMC Biol 8:109.
38. Pinkerton M, Steinrauf LK, Dawkins P (1969) The molecular structure and some
14. Currie CR, et al. (2003) Ancient tripartite coevolution in the attme ant-microbe
transport properties of valinomycin. Biochem Biophys Res Commun 35:512-518.
symbiosis. Science 299:386-388.
39. Watanabe H, Azuma M, Igarashi K, Ooshima H (2005) Valinomycin affects the
15. Santos AV, Dillon RJ, Dillon VM, Reynolds SE, Samuels Rl (2004) Ocurrence of the
morphology of Candida albicans. J Antibiot (Tokyo) 58:753-758.
antibiotic producing bacterium Burkholderia sp. in colonies of the leaf-cutting ant
40. Lim TH, et al. (2007) Antifungal activity of valinomycin, a cyclodepsipeptide from
Atta sexdens rubropilosa. FEMS Microbiol Lett 239:319-323.
Streptomyces padanus TH-04. Nat Prod Sci 13:144-147.
16. Kost C, et al. (2007) Non-specific association between filamentous bacteria and
41. Park CN, Lee JM, Lee D, Kim BS (2008) Antifungal activity of valinomycin, a peptide
fungus-growing ants. Naturwissenschaften 94:821-828.
antibiotic produced by Streptomyces sp. Strain M10 antagonistic to Botrytis cinerea.
17. Pinto-Tomás AA, et al. (2009) Symbiotic nitrogen fixation in the fungus gardens
J Microbiol of
Biotechnol 18:880-884.
leaf-cutter ants. Science 326:1120-1123.
42. Heisey RM, et al. (1988) Production of valinomycin, an msecticidal antibiotic, by
18. Sen R, et al. (2009) Generalized antifungal activity and 454-screening of Streptomyces griseus var. flexipertum var. nov. J Agrie Food Chem 36:1283-1286.
Pseudonocardia and Amycolatopsis bacteria in nests of fungus-growing ants. Proc 43. Waksman SA, Tishler M (1942) The chemical nature ot actinomycin, an antimicrobial
Nati Acad Sci USA 1 06: 1 7805-1 781 0.
substance produced by Actinomyces antibioticus. J Biol Chem 142:519-528.
19. Jensen PR, Williams PG, Oh DC, Zeigler L, Fenical W (2007) Species-specific secondary
44. Moore GÈ, Dipaolo JA, Kondo T (1 958) The chemotherapeutic effects and complications
metabolite production in marine actinomycetes of the genus Salinispora. Appi
of actinomycin D in patients with advanced cancer. Cancer 1 1:1204-1214.
Environ Microbiol 73:1 146-1 1 52.
45. Kroiss J, et al. (2010) Symbiotic Streptomycetes provide antibiotic combination
20. Matter AM, Hoot SB, Anderson PD, Neves SS, Cheng YQ (2009) Valinomycin prophylaxis for wasp offspring. Nat Chem Biol 6:261-263.
biosynthetic gene cluster in Streptomyces: Conservation, ecology and evolution. PLoS
46. Fernandez-Marin H, Zimmerman JK, Nash DR, Boomsma JJ, Wcislo WT (2009) Reduced
ONE 4:e7194.
biological control and enhanced chemical pest management in the evolution of
21. Jorgensen H, et al. (2009) Candicidin biosynthesis gene cluster is widely distributed
fungus farming in ants. Proc R Soc Lond B Biol Sci 276:2263-2269.
among Streptomyces spp. isolated from the sediments and the neuston 47. layer of the
Anukool U, Gaze WH, Wellington EMH (2004) In situ monitoring of streptothricm
Trondheim fjord, Norway. Appi Environ Microbiol 75:3296-3303. production by Streptomyces rochei F20 in soil and rhizosphere. Appi Environ
22. Cafaro MJ, Currie CR (2005) Phylogenetic analysis of mutualistic filamentous bacteria
Microbiol 70:5222-5228.
associated with fungus-growing ants. Can J Microbiol 51:441-446. 48. Kieser T, Bibb MJ, Buttner MJ, Chater KF, Hopwood DA (2000) Practical Streptomy
23. Mueller UG, Dash D, Rabeling C, Rodrigues A (2008) Coevolution between attine ants(Crowes, Norwich, United Kingdom).
Genetics
and actinomycete bacteria: A réévaluation. Evolution 62:2894-2912. 49. DeSantis TZ, et al. (2006) Greengenes, a chimera-checked 16S rRNA gene databa
24. Zucchi TD, Guidolin AS, Cònsoli FL (2010) Isolation and characterization of and workbench compatible with ARB. Appi Environ Microbiol 72:5069-5072.
actinobacteria ectosymbionts from Acromyrmex subterraneus brunneus 50. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary geneti
(Hymenoptera, Formicidae). Microbiol Res 166:68-76. analysis (MEGA) software version 4.0. Mol Biol Evol 24:1596-1599.
25. Mordarski M, Skurska H, Morawiecki A (1958) Antibacterial properties of 51. Beavis RC, Chaudhary T, Chait BT (1992) a-Cyano^-hydroxycinnamic acid as a matr
Streptomyces V. Identification of a Streptomyces parvus antibiotic. Arch Immunologi! for matrix-assisted laser desorption mass-spectrometry. Org Mass Spectrom 2
I Terapii Doswiadczalnej 6:367-373. 156-158.

1960 | www.pnas.org/cgi/doi/10.1073/pnas.1008441108 Schoenian et al.

This content downloaded from 14.139.85.2 on Fri, 20 Jul 2018 12:03:16 UTC
All use subject to http://about.jstor.org/terms